Life looks increasingly like a chemical experiment that took over the laboratory. All living things turn to dust and ashes when they die, or, to put it another way, to constituent atoms and molecules of hydrogen, oxygen, carbon, phosphorus and so on.
But, in another sense, living things do not die: they begin again, from a tiny cell, and scavenge the dust, the air and water, to find the elements necessary to fashion an aspidistra, an elephant, or an attorney-general, using only the raw materials to hand and energy from a thermonuclear reactor 93 million miles away. The freshly minted, self-replicating organism then grows up, grows old and melts away, but not before imparting a fragment of itself to generate yet another copy, but not an identical copy. The process is visible and transparent, everywhere on the planet, but it is ultimately mysterious. It has been going on for at least 3.5bn years, but researchers may never satisfactorily explain how it ever got started.
The mystery may endure because, once up and running, the life machine kicked up enough dust to cover its original tracks. It altered the air, muddied the water and recycled the rocks around it. For at least 30 centuries, thinkers ascribed the beginning of life to an extraterrestrial agency: they talked of the hand of God, the divine afflatus, the vital spark, or of "seeds" of life travelling through the cosmos. By 1850, however, chemists, physicists, geologists and biologists - many of them deeply religious, and all of them familiar with the religious tradition - had begun to take the problem seriously, and concluded that complex life had in some sense evolved from simpler beginnings, exploiting the materials around it for its own survival.
Charles Darwin in 1859 proposed that life may have brewed in a soup of organic chemicals in some "warm little pond" on the surface of the primordial Earth. He left the question open: it remains open. His inheritors have proposed that life could have been generated in the first sunlit oceans that swept across the young planet, or in the crater left by an asteroid impact, or made a template of itself in a bed of wet clay, or in the dark silence of submarine volcanoes, or that it had been delivered in a meteoritic fragment from a faraway planet.
But as far as we know, life exists on Earth and nowhere else. This is a puzzle because, at one level, the universe looks as though it was set up to generate life. In the first place, the constants of physics are so finely tuned that, were they even infinitesimally different, there would be no stars or planets, no carbon atoms, no oxygen, no aspidistras, elephants or attorney-generals. In the second place, the space between the stars - where no life could ever exist - is rich in life's prime ingredients: the organic chemicals. Astronomers have identified more than 100 of these, including cyanide, formaldehyde, alcohol, ammonia, and acetylene. Comets are rich in hydrocarbons. A meteorite that fell to Earth in Australia in 1969 has so far yielded more than 70 amino acids. These are the building blocks of protein. In 1953, two Chicago chemists filled a flask with ammonia, methane, water and hydrogen - the Earth's primitive atmosphere must have contained all these - and ran an electric current through it. After just a week, they had 13 of the 22 amino acids that are the constituents of protein, the stuff of all living material. If that happened in one week in one laboratory, they reasoned, it could certainly have happened through a billion years of lightning strikes.
But it is a big jump from life's building blocks to self-replicating, planet-altering life, so big that the astronomer Sir Fred Hoyle famously argued that life had as much chance of emerging from a chemistry set by accident, as a jumbo jet had of appearing from a windstorm in a junkyard. Yet life exists, it shares a common biochemistry, and species are grouped in such a way as to suggest common ancestry. But how it began, and precisely when and where, remains a mystery.
Life looks after itself, but humans look after the classification of life, and humans argue. The first biologists divided life into animal and plant kingdoms, and then subdivided these into phyla, or divisions, classes, orders, families, genera, species and subspecies. But this was long before the discovery of DNA, which showed that creatures that seemed similar could nevertheless have evolved from very different lineages, and so categories are always under revision. The process of classification began long before anyone realised that microbes and their parasitic viruses dominate all life on the planet. Complex life was a late arrival. This microbial influence is so profound that another biologist, Carl Woese, proposed just three kingdoms: archaea, bacteria and eucaryota. All of these are family trees with many branches, but the last group, the microbes with nuclei, end with three little twigs from which evolved all the planet's plants, animals and fungi. Biologists now argue the case for either five (monera, protista, plantae, fungi, animalia) or six kingdoms (plants, animals, protists, fungi, archaebacteria and eubacteria).